Range Loss Over Time, What Can Be Expected, Efficiency, How to Maintain Battery Health

[Jeremy3292]

I just spent about an hour reading the entire document lol (nerd). My takeaway was there was a ton of data in there but out of the things you can control you should basically just keep it between 20 and 80% and you’ll keep 90%+ capacity over a long period of time. If you really don’t drive a lot you can try and keep the SOC between 40 to 60% for a marginal benefit but it is almost statistically insignificant compared to 20 to 80%

 

[Aphelion]

So the sweet spot is around 80%.

 

[Candleflame]

Well, it's "OK" to keep it at 90%, but the data shows lower SOCs is better for less degradation.

what i dont get about this is that with a standard lithium ion battery 100% at 15C is the same as i.e. 80% at 35C. By that logic, us in Australia or i.e. Texan Model 3s should have extremely heavy degradaton. But that is not the case.

 

[Jeremy3292]

So the sweet spot is around 80%.

The sweet spot is technically 40-60%, but keeping it between 20-80% appears to have no statistically significant impact vs 40-60%.

 

[Jeremy3292]

what i dont get about this is that with a standard lithium ion battery 100% at 15C is the same as i.e. 80% at 35C. By that logic, us in Australia or i.e. Texan Model 3s should have extremely heavy degradaton. But that is not the case.

I don't follow. Which graph are you referring to?

 

[Candleflame]

I don't follow. Which graph are you referring to?

theres a graph which shows the battery stress of temperature vs SOC. need to have a look around to find it.
Essentially a 10C 100% SOC the stress on the battery seems quite similar as i.e. 90% at 25C.

 

[AAKEE]

I try not to keep it at 90% for a long time either. I will charge it to 90% at work sometimes but then I’ll use it within a few hours to get home. So when I’m at home it’s never higher than 80% sitting over night. I like to use the chargers at work as much as I can as they are free. My understanding is it is fine if you charge it to 90% (or even 100%) as long as you use it within a few hours and not let it sit for a while.

i’ve never really understood the shallower discharge theory. So if you use 10% of the battery you should charge it immediately versus using 10% each day over three days and then charging?

The lower the average SOC, the lower the degradation from time( ”calendar aging” )

In many cases calendar aging is the main degradation. Its the use of the car that decides if it is the cyclic aging or the calendar aging that causes the mist degradation.
Low SOC and smaller Depth Of Discharge(DOD) is better for the battery longevity.

The DOD is the discharge in percent, ie after carged to 80% if discharged to 30%, the DOD is 50%. For the same DOD, its better with a lower span, for example 60-30% is better than 80-50%, even if both DOD is 30%.

Its better to charge daily 15% than to wait three days and charge 45%.

BTW: The shallow discharge theory is not a ”theory”. Its a fact that is supported by a lot of research reports.

 

[Jeremy3292]

The lower the average SOC, the lower the degradation from time( ”calendar aging” )

In many cases calendar aging is the main degradation. Its the use of the car that decides if it is the cyclic aging or the calendar aging that causes the mist degradation.
Low SOC and smaller Depth Of Discharge(DOD) is better for the battery longevity.

The DOD is the discharge in percent, ie after carged to 80% if discharged to 30%, the DOD is 50%. For the same DOD, its better with a lower span, for example 60-30% is better than 80-50%, even if both DOD is 30%.

Its better to charge daily 15% than to wait three days and charge 45%.

BTW: The shallow discharge theory is not a ”theory”. Its a fact that is supported by a lot of research reports.

So if you use 10% charge daily it would be better to go from 70% to 60% and then charge back to 70% daily then to go from 70% to 40% (in total over 3 days) and then charge back to 70% and repeat that same cycle every 3 days.

 

[AAKEE]

I just spent about an hour reading the entire document lol (nerd). My takeaway was there was a ton of data in there but out of the things you can control you should basically just keep it between 20 and 80% and you’ll keep 90%+ capacity over a long period of time. If you really don’t drive a lot you can try and keep the SOC between 40 to 60% for a marginal benefit but it is almost statistically insignificant compared to 20 to 80%

The lower the SOC, the better the it is for the battery. This is the case both when the car sleeps and when it is used.
The 20% comes from Tesla än is not a question of the degradation of the li ion battery. They degrade less the lower SOC you have. But the car has a 12V lead battery that both doesnt like being discharged at all and there will be problems for the owner if it gets completely discgarged.
Theres a clear benifit of storing a NCA battery below some 57% SOC. The degradation from time will be about half, and also if cycling is done with smaller DOD and at lower SOC the aging from cycling will be much smaller. Also, keeping the SOC low on really hot days is a good idea.
This above is most certain the reason why my ’21 M3P with the 82kWh batttery still shows full range(507-508km) after 22000km/ 9 months of use. My nominal full pack is at 80.9kWh today with a starting point of 80.7.
Down below from a tesearch report of a Panasonic NCA lion( 18650, should be the same or really close to the S/C cells.

 

[AAKEE]

So if you use 10% charge daily it would be better to go from 70% to 60% and then charge back to 70% daily then to go from 70% to 40% (in total over 3 days) and then charge back to 70% and repeat that same cycle every 3 days.

Exactly and absolutely.
Theres a reason Tesla say ”connect the car to the charger whenever you can”.

Only reason not to charge daily is if you dont have an own charging point so you need to travel for the charging( time loss etc).

 

[Jeremy3292]

This above is most certain the reason why my ’21 M3P with the 82kWh batttery still shows full range(507-508km) after 22000km/ 9 months of use. My nominal full pack is at 80.9kWh today with a starting point of 80.7.

I believe this is actually heavily disputed in this thread as there is a lot of evidence that the 82 kWh battery's full potential is "software locked" and is not being utilized fully and thus will not show degradation for awhile.

@eivissa has extensively wrote about this. I'm sure my wording is improper.

 

[AAKEE]

The sweet spot is technically 40-60%, but keeping it between 20-80% appears to have no statistically significant impact vs 40-60%.

Id say that the sweet spot is as low as possible when it comes to degradation. Teslas minimum charging level is 50% so you wont get below that if you dont interrupt the charging. 50% is low enough, and below the 57-60% SOC or more causing more calendar aging.

 

[Jeremy3292]

Id say that the sweet spot is as low as possible when it comes to degradation. Teslas minimum charging level is 50% so you wont get below that if you dont interrupt the charging. 50% is low enough, and below the 57-60% SOC or more causing more calendar aging.

I was going off this graph attached above at right. You can see the mean % capacity of 20-80% SOC range and 40-60% SOC range are virtually identical based on the Y axis. Looks like they are 95% and 97% based on my squinty eye reading.

 

[AAKEE]

I believe this is actually heavily disputed in this thread as there is a lot of evidence that the 82 kWh battery's full potential is "software locked" and is not being utilized fully and thus will not show degradation for awhile.

@eivissa has extensively wrote about this. I'm sure my wording is improper.

The 82kWh in the M3P is not software locked.
I use Scan my Tesla and can see that a full charge put the cells to 4.20V/ cell, which is the standard for 100% SOC.
Also, I might have the record for the 82.kWh battery with a nominal full of 81.6kWh, at that time with 82.0kWh remaining

I know eivissa also got a good number but I think he was slightly lower

As for the 82kWh in Europe, Im not sure its software locked either. I havent followed the case lately but I know the smaller 77.8 kwh battery was software locked to make similar range as the smaller 75kWh from LG.

Back to the 82kWh battery the issue seems more to be that people get smaller capacity than 82kwh, and quite different onces, from 77.X to 80kWh.

 

[eivissa]

I believe this is actually heavily disputed in this thread as there is a lot of evidence that the 82 kWh battery's full potential is "software locked" and is not being utilized fully and thus will not show degradation for awhile.

@eivissa has extensively wrote about this. I'm sure my wording is improper.

Well the 82kWh pack is optically locked in the LR (E3LD) to 77,8kWh. Any capacity above will not show and the range is capped at 567-575k.

The 82kWh pack in the Performance is also optically locked, but higher in the region of 80,6-80,9kWh. Any capacity above will also not show and the car is capped at 508-509km.

None of the two are actually software locked. If treated well and calibrated well they will both go to 82kWh.

 

[AAKEE]

View attachment 708862

I was going off this graph attached above at right. You can see the mean % capacity of 20-80% SOC range and 40-60% SOC range are virtually identical based on the Y axis. Looks like they are 95% and 97% based on my squinty eye reading.

What you do not see in that graph is the difference between the same DOD but with different starting SOC, the 60-40% has a DOD of 20%.
If you instead use it from 40-20% you will get even better result, or actually 30-10% is even better.
You might need to read a lot of research reports as it both sometimes come out the wrong way in one report. This is both from maybe not setting up the test in the best way, which in turn make you draw the wrong conclusions. Also, as most cases, the result is valid and the conclusions also but the reader might draw the wrong conclusions, as it most probably is in this case.
The rrlort is right but it doesnt show the whole truth.

 

[Jeremy3292]

The 82kWh in the M3P is not software locked.
I use Scan my Tesla and can see that a full charge put the cells to 4.20V/ cell, which is the standard for 100% SOC.
Also, I might have the record for the 82.kWh battery with a nominal full of 81.6kWh, at that time with 82.0kWh remaining

I know eivissa also got a good number but I think he was slightly lower

As for the 82kWh in Europe, Im not sure its software locked either. I havent followed the case lately but I know the smaller 77.8 kwh battery was software locked to make similar range as the smaller 75kWh from LG.

Back to the 82kWh battery the issue seems more to be that people get smaller capacity than 82kwh, and quite different onces, from 77.X to 80kWh.

Thanks for the insight. I am a novice but I do enjoy a good technical reading. I have free charging at work and utilize 10-12% each way on my commute. My plan is:

• Charge to 70% at work
• Arrive back home around 58-60%, car sleeps at this SOC
• Arrive at work next day around 46-48%, charge to 70%
• Repeat cycle

I have a HPWC at home and will utilize it on the weekends or days where I'm not in the office. I figure I might as well utilize the free charging at work and only pay to charge at home when I need to!

 

[AAKEE]

Well the 82kWh pack is optically locked in the LR (E3LD) to 77,8kWh. Any capacity above will not show and the range is capped at 567-575k.

The 82kWh pack in the Performance is also optically locked, but higher in the region of 80,6-80,9kWh. Any capacity above will also not show and the car is capped at 508-509km.

None of the two are actually software locked. If treated well and calibrated well they will both go to 82kWh..

Yes, the bold part is important, as some might not see the difference of “locked” in the energy graph [backwards calculation] and a real capacity lock odf the battery.

 

[Jeremy3292]

What you do not see in that graph is the difference between the same DOD but with different starting SOC, the 60-40% has a DOD of 20%.
If you instead use it from 40-20% you will get even better result, or actually 30-10% is even better.
You might need to read a lot of research reports as it both sometimes come out the wrong way in one report. This is both from maybe not setting up the test in the best way, which in turn make you draw the wrong conclusions. Also, as most cases, the result is valid and the conclusions also but the reader might draw the wrong conclusions, as it most probably is in this case.
The rrlort is right but it doesnt show the whole truth.

Lower SOC is better but for a Tesla you don't want to get below 20% bc that may kick the 12v battery into action? I believe you mentioned something like that earlier.

 

[AlanSubie4Life]

I do not hink you need to loose 10% over two years if you can adhere to the basics of battery degradation and follow that knowledge

I'm definitely on board with the strategy of charging to ~60% or so and keeping depth of discharge to a minimum. It's the right choice according to the evidence (if it works for the end user - it doesn't work for everyone due to particular (common) use cases).

But whether you "need to lose 10% over two years" - that is definitely the big question! I haven’t seen a ton of evidence that specific optimal treatment of the battery is strongly correlated to capacity loss for Tesla vehicles (even though it should be!), but sticking to proven best practices for Li-ion batteries can’t hurt, and probably does help in general. Certainly there are a couple examples here of people using the lower SOC and having good results. But overall trends? I have no idea. It seems like there are plenty of examples of people with ~70% SOC habits with low depth of discharge daily charging, which you’d think would be pretty healthy, who see ~10% capacity loss after a couple years. And others with similar habits who are doing noticeably better.

Time will tell for 2021 vehicles. Was interesting to see the TeslaFi fleet data for 2021 - but a very small sample size it sounds like? And possibly some TeslaFi users would be predisposed to higher capacity loss (might be one of the reasons they sign up, for example!).

Will see what the next year brings…

If treated well and calibrated well they will both go to 82kWh.

I agree they don't seem to be software locked.

I know that some LR vehicles exist which have very healthy packs...close to 82kWh (though I think I'm taking your word for this (which I believe) and you've only seen them in other forums - I don't remember a direct report here? - my memory is poor as discussed previously on this topic, haha).

However, whether MOST LR vehicles will go to those levels is (I think) subject to some debate. From the reports I've seen, we routinely see 79-80kWh values for these LRs, and seldom see higher values, even with various 60-70% charging strategies employed.

Now...it true that's probably not a software lock, if there is a difference. But just searching for an explanation of what (I think) seems to be a small discrepancy in NFP between LRs (on average) and Ps (on average). Does that difference truly exist? Not sure.